Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Short Exposure Time ASI 1600MM Cool?

This topic has been archived. This means that you cannot reply to this topic.
36 replies to this topic

#1 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 21 May 2017 - 09:05 PM

Okay, I ran some tests the other night on M81-M82 using the following recommended exposure settings:

 

Optimal Exposure:
Median ADU shown in SGP -
Gain 0 Offset 10:         400 ADU
Gain 75 Offset 12:       550 ADU
Gain 139 Offset 21:     850 ADU
Gain 200 Offset 50:    1690 ADU
Gain 300  Offset 50 :  2650 ADU
Better to be a little high than low
For other Gains or offsets use the formula Jon indicated: MinDN16 = (((ReadNoise * 20) / Gain) + BiasOffset) * 2^16/2^Bits
Dithering:
Just pick "X" for dithering every "X" frames to be 5% of the number of subs you are going to collect. If you are collecting 20 subs, you should dither every frame. If you are collecting 50 subs, dither every 2 frames (OK the rule should say "round down"), etc.
Calibration Frames:
             Flats:
                      In SGP you want your flats to measure 12k-18k DN.
             Bias:
                      Use Dark Flats instead of Bias if Flats long.
Get new calibration frames when changing driver.

 

See link https://www.cloudyni...-sheet-no-math/

 

The attached image is a cropped calibrated (dark, bias and flat at 15k) 1 minute sub in a fairly light polluted sky. I use the IDAS D2 LPS. The image was taken at a Gain of 0 and Offset of 10. The Fit image stats were as follows:

 

Mean: 685

Median: 684

Min: 125

Maximum: 64461

 

The center of M81 reads as follows:

Mean: 2348
Median: 2014
Min: 1165
Maximum: 6174

 

The bright center of M82 reads as follows:

 

Mean: 1163
Median: 1132
Min: 782
Maximum: 1853

 

Following the advice of that thread, the targets M81 and M82 seem woefully underexposed. A stack of ~150 subs did not yield a useful picture as I had to stretch and stretch the image to bring out the galaxy arms and the BG by that point looked horrible.

 

I think I understand the proposed concept of keeping the BG at about 20x the read noise or higher so as to not clip too much data or reducing the dynamic range. But it seems that this theory breaks down in actual practice. It seems to me that clipping star data cannot be avoided if much of the target details are going to be brought out. If I increased the exposure time perhaps by 15-30 seconds, the star data starts getting clipped. I am not sure it is a bad thing to clip star data if you are not going to be able to reasonably bring out target details. What is the harm of clipping star data? I sure it is possible that I am doing something wrong here. If the thread exposure time is correct, It would seem to me that I would need upwards of 500 or more subs to get the target detail out and I tend to doubt that would help if a stack of 150 was unusable.

 

It also seems to me that the target stats are way too low to bring out much information even though the calibrated BG is a little higher than the thread recommends. Again, I may be doing something wrong or not understand this exposure thing. I am wondering if a 3 minute sub would yield much better results even though the BG stats would be in the 3-5k at this gain setting (0 and 10).

 

I am probably opening up a can of worms here, but I am trying to learn this camera and my new EON 115 scope. 

 

If you guys have recommendations as to what exposure to use without blowing the BG out, that would be helpful as well.

 

I know Jon Rista can probably tell me where I am going wrong as I know he likes the short exposure times.

 

Thanks

 

Kaos

Attached Thumbnails

  • A M81-M82.jpg

Edited by Kaos, 21 May 2017 - 09:30 PM.


#2 Alien Observatory

Alien Observatory

    Vanguard

  • *****
  • Posts: 2,197
  • Joined: 13 Nov 2015

Posted 21 May 2017 - 09:27 PM

Kaos, Just a thought in reference to your problem.  It seem's to me the EON 115 at F7 is pushing the boundaries for lack of photons.  It is a small scope (4.5") with a relativity large F ratio (F7).  

 

That being said some type of focal reduction may be a step in the right direction to get more photons to the camera sensor.  Longer subs and/or more of them will also be a step in the right direction.

 

Sorry not a lot of help...Pat Utah



#3 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 21 May 2017 - 10:05 PM

Kaos, the main missing piece of the puzzle for you is total integration time. While this camera is better used with short subs, you still need enough total integration time. At 150 subs x 1 minute, you have 150 minutes of data. That is two and a half hours, a decent amount of time, but if you really want to go deeper, you need more than that. To double your exposure depth, you need 600 subs. 

 

Now, that is a lot of subs. You should consider using longer exposures. Given you are using Gain 0, which has a 20ke- FWC, AND considering you are using an LPS-D1, AND considering you are using an f/7 scope, you should be able to use longer exposures than that. I use an f/4 scope in a red zone, and I can use 60 second subs without any filters in ~18.2mag/sq" skies. Assuming you have similar skies, you have A) the LPS-D1 which should cut out about half the light, so that is a 1 stop difference, and B) you have an f/7 scope which is another 1.75 stops. That is ~6.72x difference in exposure (assuming your skies are indeed the same as mine), which means you should be able to expose a good deal longer than you have been.

 

I would say definitely give the 3 minute subs a try, and I would be surprised if you really were clipping much data with such exposures. You might even be able to use longer subs, however if that is the case, my recommendation is to try a higher gain and stick to around 3 minute subs, because in the end you will want to stack at least 100 subs, and 250 subs is more ideal. If you get 150x3m subs, that is 7.5 hours, and you should have a pretty nice image at that point (plus, remember, you will want some RGB time as well, so your total integration time would likely top 10 hours, which would be great!)



#4 petert913

petert913

    Fly Me to the Moon

  • *****
  • Posts: 5,747
  • Joined: 27 May 2013

Posted 21 May 2017 - 10:08 PM

Yes,  here are my scant  6x60 second frames taken with my 80mm Stellarvue refractor and a Canon DSLR.  It seems to show much more luminosity.

That's only 6 minutes total......

 

M81-M82_final_crop_zps033msyx1.jpg


Edited by petert913, 21 May 2017 - 10:12 PM.


#5 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 21 May 2017 - 10:21 PM

Kaos, the main missing piece of the puzzle for you is total integration time. While this camera is better used with short subs, you still need enough total integration time. At 150 subs x 1 minute, you have 150 minutes of data. That is two and a half hours, a decent amount of time, but if you really want to go deeper, you need more than that. To double your exposure depth, you need 600 subs. 

 

Now, that is a lot of subs. You should consider using longer exposures. Given you are using Gain 0, which has a 20ke- FWC, AND considering you are using an LPS-D1, AND considering you are using an f/7 scope, you should be able to use longer exposures than that. I use an f/4 scope in a red zone, and I can use 60 second subs without any filters in ~18.2mag/sq" skies. Assuming you have similar skies, you have A) the LPS-D1 which should cut out about half the light, so that is a 1 stop difference, and B) you have an f/7 scope which is another 1.75 stops. That is ~6.72x difference in exposure (assuming your skies are indeed the same as mine), which means you should be able to expose a good deal longer than you have been.

 

I would say definitely give the 3 minute subs a try, and I would be surprised if you really were clipping much data with such exposures. You might even be able to use longer subs, however if that is the case, my recommendation is to try a higher gain and stick to around 3 minute subs, because in the end you will want to stack at least 100 subs, and 250 subs is more ideal. If you get 150x3m subs, that is 7.5 hours, and you should have a pretty nice image at that point (plus, remember, you will want some RGB time as well, so your total integration time would likely top 10 hours, which would be great!)

Thanks Jon... I was a little confused by such a low BG recommendation in the thread I referred to. If I am understanding you correctly, if I used 3 min subs for example, and the median was 5K, you would not say that is contrary to the thread recommendations? The 400 level at that gain setting seemed low to me, but you guys are the ones in the know! Also, if I understand you correctly, assuming my skies are the same as yours (I am in a burgundy zone bordering an orange zone), I could expose 6.72 times your exposure time with no filters on your end?

 

Thanks

 

Kaos



#6 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 21 May 2017 - 10:33 PM

 

Kaos, the main missing piece of the puzzle for you is total integration time. While this camera is better used with short subs, you still need enough total integration time. At 150 subs x 1 minute, you have 150 minutes of data. That is two and a half hours, a decent amount of time, but if you really want to go deeper, you need more than that. To double your exposure depth, you need 600 subs. 

 

Now, that is a lot of subs. You should consider using longer exposures. Given you are using Gain 0, which has a 20ke- FWC, AND considering you are using an LPS-D1, AND considering you are using an f/7 scope, you should be able to use longer exposures than that. I use an f/4 scope in a red zone, and I can use 60 second subs without any filters in ~18.2mag/sq" skies. Assuming you have similar skies, you have A) the LPS-D1 which should cut out about half the light, so that is a 1 stop difference, and B) you have an f/7 scope which is another 1.75 stops. That is ~6.72x difference in exposure (assuming your skies are indeed the same as mine), which means you should be able to expose a good deal longer than you have been.

 

I would say definitely give the 3 minute subs a try, and I would be surprised if you really were clipping much data with such exposures. You might even be able to use longer subs, however if that is the case, my recommendation is to try a higher gain and stick to around 3 minute subs, because in the end you will want to stack at least 100 subs, and 250 subs is more ideal. If you get 150x3m subs, that is 7.5 hours, and you should have a pretty nice image at that point (plus, remember, you will want some RGB time as well, so your total integration time would likely top 10 hours, which would be great!)

Thanks Jon... I was a little confused by such a low BG recommendation in the thread I referred to. If I am understanding you correctly, if I used 3 min subs for example, and the median was 5K, you would not say that is contrary to the thread recommendations? The 400 level at that gain setting seemed low to me, but you guys are the ones in the know! Also, if I understand you correctly, assuming my skies are the same as yours (I am in a burgundy zone bordering an orange zone), I could expose 6.72 times your exposure time with no filters on your end?

 

Thanks

 

Kaos

 

It is contrary...however, there is a reasonable number of subs to stack. Stacking 600 1-minute subs is pretty crazy. ;) I don't really like to stack much more than 300 subs per channel myself, and I "break the rules" in order to achieve that. In my case, with my scope, I actually don't even "need" 60 second L subs. Such subs are much more exposed than I need, and will give me a 2000 ADU background sky.

 

That said, a 5000 16-bit ADU background sky is really high. Are you SURE you will get that in 3 minutes with an f/7 scope and an IDAS LPS-D1? If that is indeed the case, then your skies are significantly worse than mine. Significantly worse...like, dead smack in the middle of a white zone, downtown metropolitan city. If that is more the case, then I highly recommend you ditch the LRGB, pick up a set of Ha, OIII and SII filters, and stick to narrow band imaging. You are otherwise fighting so much LP that you will need a lot more than 10 hours to get reasonable results, and in the end, it is really just not worth the effort (and believe me, I've spent probably 250 hours or so acquiring 15-20 hours of data on objects that I just couldn't create good images out of...all because of the LP...) 

 

Either switch to narrow band, or find a dark site you can use on a regular basis, even if it is a yellow zone. You will get good results in just a few hours, and they will completely blow you away in comparison to what you can get from your light polluted site. 



#7 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 21 May 2017 - 10:43 PM

 

 

Kaos, the main missing piece of the puzzle for you is total integration time. While this camera is better used with short subs, you still need enough total integration time. At 150 subs x 1 minute, you have 150 minutes of data. That is two and a half hours, a decent amount of time, but if you really want to go deeper, you need more than that. To double your exposure depth, you need 600 subs. 

 

Now, that is a lot of subs. You should consider using longer exposures. Given you are using Gain 0, which has a 20ke- FWC, AND considering you are using an LPS-D1, AND considering you are using an f/7 scope, you should be able to use longer exposures than that. I use an f/4 scope in a red zone, and I can use 60 second subs without any filters in ~18.2mag/sq" skies. Assuming you have similar skies, you have A) the LPS-D1 which should cut out about half the light, so that is a 1 stop difference, and B) you have an f/7 scope which is another 1.75 stops. That is ~6.72x difference in exposure (assuming your skies are indeed the same as mine), which means you should be able to expose a good deal longer than you have been.

 

I would say definitely give the 3 minute subs a try, and I would be surprised if you really were clipping much data with such exposures. You might even be able to use longer subs, however if that is the case, my recommendation is to try a higher gain and stick to around 3 minute subs, because in the end you will want to stack at least 100 subs, and 250 subs is more ideal. If you get 150x3m subs, that is 7.5 hours, and you should have a pretty nice image at that point (plus, remember, you will want some RGB time as well, so your total integration time would likely top 10 hours, which would be great!)

Thanks Jon... I was a little confused by such a low BG recommendation in the thread I referred to. If I am understanding you correctly, if I used 3 min subs for example, and the median was 5K, you would not say that is contrary to the thread recommendations? The 400 level at that gain setting seemed low to me, but you guys are the ones in the know! Also, if I understand you correctly, assuming my skies are the same as yours (I am in a burgundy zone bordering an orange zone), I could expose 6.72 times your exposure time with no filters on your end?

 

Thanks

 

Kaos

 

It is contrary...however, there is a reasonable number of subs to stack. Stacking 600 1-minute subs is pretty crazy. wink.gif I don't really like to stack much more than 300 subs per channel myself, and I "break the rules" in order to achieve that. In my case, with my scope, I actually don't even "need" 60 second L subs. Such subs are much more exposed than I need, and will give me a 2000 ADU background sky.

 

That said, a 5000 16-bit ADU background sky is really high. Are you SURE you will get that in 3 minutes with an f/7 scope and an IDAS LPS-D1? If that is indeed the case, then your skies are significantly worse than mine. Significantly worse...like, dead smack in the middle of a white zone, downtown metropolitan city. If that is more the case, then I highly recommend you ditch the LRGB, pick up a set of Ha, OIII and SII filters, and stick to narrow band imaging. You are otherwise fighting so much LP that you will need a lot more than 10 hours to get reasonable results, and in the end, it is really just not worth the effort (and believe me, I've spent probably 250 hours or so acquiring 15-20 hours of data on objects that I just couldn't create good images out of...all because of the LP...) 

 

Either switch to narrow band, or find a dark site you can use on a regular basis, even if it is a yellow zone. You will get good results in just a few hours, and they will completely blow you away in comparison to what you can get from your light polluted site. 

 

I am estimating the 3-5k based on close to 2K I got in testing a 90 sec sub the other night. Maybe that is on the high side. Plus, from by observatory location, it is significantly brighter imaging north and much darker in the other three directions. North is in the direction of a gray zone located about 15-20 miles away. Okay, I will give 3-5 min subs a shot and see how they turn out. Good to know that you guys in the know with this camera "break the rules" lol. So, as long as I am not really washing out the BG and not clipping any data on the target I should be okay?

 

Kaos



#8 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 21 May 2017 - 11:55 PM

If you are shooting through the LP bubble over the center of a downtown area, that can be almost as bad as being right inside it...and even possibly worse (since you would be shooting at an angle, so you could be shooting through a larger heavily light polluted airmass than if you were pointing strait up through that same LP bubble.) 

 

In my honest opinion, if you are shooting through white zone LP for LRGB, then you are most likely wasting time. You would need far more than 10 hours of data to get a decent result...probably more like 20-30 hours. I've integrated about 20 hours on a few targets, back when I was measuring 18.5mag/sq" overhead, and I was never able to make anything out of those images. LP is a real RGB/LRGB killer, and it's bad enough in a red zone without shooting through a white zone dome. 

 

I'd either find targets in directions where the sky is darker, at the very least, or find a dark site, or switch to NB. With NB (more nebula targets rolling around soon), you can get significantly better data in a lot less time. Most of my ASI1600 NB images are around 3 hours a channel (and often just the Ha channel due to limited clear sky time). The data is vastly superior to any of my LRGB data. 



#9 suvowner

suvowner

    Viking 1

  • *****
  • Posts: 662
  • Joined: 22 May 2016

Posted 25 May 2017 - 01:13 AM

check out this blog post by Sam

 

https://astronomy-im...n-ascom-driver/



#10 vdb

vdb

    Surveyor 1

  • -----
  • Posts: 1,618
  • Joined: 08 Dec 2009

Posted 25 May 2017 - 02:53 AM

OK, look at my thread:

https://www.cloudyni...p-zone-ngc6946/

 

It was a dry night, but my location is not great, and NGC6946 passed at the north of me, I included an LP map.

 

I took all 30 sec subs gain 30, offset 60 QHY163M and to be frankly a single sub was horrible, default screen stretch + statistics, you see just take lots of them and stretch the hell out of the stack you can.

My background was around 9000 with 30 sec subs ...

 

EDIT:

my RGB subs are around 2000 background ...

Attached Thumbnails

  • Screen Shot 2017-05-25 at 09.46.58.jpg

Edited by vdb, 25 May 2017 - 03:17 AM.


#11 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 25 May 2017 - 06:22 PM

check out this blog post by Sam

 

https://astronomy-im...n-ascom-driver/

Thanks for the link SUV. I think those tests are pretty revealing. The brighter the target, the lower the gain setting. It is also pretty obvious that you get much better star (for double and triples) control on the HDR setting although you will have to have more exposures to bring out the same faint detail on some targets. I guess it will be trial and error with each target taking a few test shots to find an acceptable exposure point. Sam does make a good point that once you get to a certain exposure level, the sky glow will cover the read noise and you get no substantial difference between the HDR setting and the LRN setting. I hadn't stop to think about that point. So, based on a wasted night of images, Jon's points above and Sam's tests, I am not sure I totally buy into the recommended exposure cheat sheet referred to above (maybe more so with a faster scope and larger aperture). I may be wrong on this, but it seems to me that the only thing the cheat sheet does is add to the number of subs you have to capture (and integration time) to gain detail when longer subs may get you there with fewer subs (less total integration time).

 

Kaos



#12 suvowner

suvowner

    Viking 1

  • *****
  • Posts: 662
  • Joined: 22 May 2016

Posted 01 June 2017 - 08:03 PM

i think the simplest way is to start with same basic 3 options in sam's blog post  and then use the histogram to make the final adjustments and then run your sequence......the nebulosity help files make this process pretty straightforward.......



#13 suvowner

suvowner

    Viking 1

  • *****
  • Posts: 662
  • Joined: 22 May 2016

Posted 01 June 2017 - 08:22 PM

from nebulosity help file

Attached Thumbnails

  • Screen Shot 2017-06-01 at 8.21.50 PM.png

Edited by suvowner, 01 June 2017 - 08:24 PM.


#14 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 01 June 2017 - 11:18 PM

from nebulosity help file

suvowmner...that is the way I use to do it until I read the thread on this camera regarding short exposure times. I managed to get a few hours of data on M101 the other night at 3 min subs. They came out good, but I do not have enough of them as they are still a bit grainy. I have 69x3mins usable subs. They look good, but I am estimating I need another ~80 subs. The BG median was around 1280. So, I think that thread might produce good results in dark skies with this camera, but I think it is woefully inadequate in light polluted skies.

 

Kaos



#15 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 02 June 2017 - 02:08 AM

 

from nebulosity help file

suvowmner...that is the way I use to do it until I read the thread on this camera regarding short exposure times. I managed to get a few hours of data on M101 the other night at 3 min subs. They came out good, but I do not have enough of them as they are still a bit grainy. I have 69x3mins usable subs. They look good, but I am estimating I need another ~80 subs. The BG median was around 1280. So, I think that thread might produce good results in dark skies with this camera, but I think it is woefully inadequate in light polluted skies.

 

Kaos

 

So with a 16-bit median around 1280, that is a 12-bit median of around 80, minus the 10 ADU bias offset so 70 ADU of true signal. At a Gain setting of 0, that is 4.88e-/ADU, which means your median background signal was 342e-. That is a darn good background signal. You'll handily swamp the read noise with that, regardless of what rule you might want to follow. With the 20x rule you would need a background signal of 70e-, and with the 3x rule you would need a background signal of only 37e-. So, fundamentally, there is no NEED to expose longer...not from an IQ standpoint. The only reason to expose longer is just to stack fewer subs...however, that is not necessarily ideal for this camera. You really want to stack 100+ subs, well dithered subs at that, to get the best results.

 

I recently imaged a group of galaxies, with 30 second subs on my f/4 scope under very bright skies at Gain 76 with an offset of 15. This gain has 2e- read noise and an ~8192e- full well. My backgrounds measured ~2600 16-bit, which is 162 12-bit. Subtracting the 15 ADU offset, I had 148 ADU, which at 2e-/ADU is 296e-. Again, I've more than sufficiently swamped the read noise, regardless of what mathematical rule you use to determine what is enough to swamp. I stacked 413 of these 30 second L subs, dithered aggressively every 4 subs, and ended up with this:

 

ly2qB44.jpg

 

Here is a closeup of NGC5371, to show the noise quality:

 

t4DI4BI.jpg

 

I think the noise looks great, which BTW is as it was in the original stack, this image so far has had zero noise reduction of any kind, only stacking and gradient removal. This camera delivers wonderful results when you stack a lot of frames. The effective real-world bit depth here is 16.4 bits. 

 

Now, the stretch is a fairly hefty one. I plate solved and annotated the image, and highlighted all the background fuzzies between 18.5 and 19.5 mag surface brightness:

 

dqogvKT.jpg

 

The faintest object highlighted in this field is 19.49mag. The faintest object found in the entire full field of the image was over 21mag! My skyfog levels measured with an SQM-L were 18.1 and 18.3 on the nights I imaged.



#16 vdb

vdb

    Surveyor 1

  • -----
  • Posts: 1,618
  • Joined: 08 Dec 2009

Posted 02 June 2017 - 03:06 AM

To complement Jon's observation;

Lot's of 30sec subs from a severe LP zone with no true dark nights anymore,

QHY 163M 280x30 sec L, 90x30secRGB/channel, GSO RC 8, AP 0.67 red, Avalon Lineair

 

34536009910_9fef73c3e0_c.jpgNGC7023 by Yves, on Flickr



#17 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 02 June 2017 - 10:27 AM

vdb, nice image. Great star color. Only suggestion is you might want to dither more aggressively. It seems you have some correlated noise in there which would be eliminated with larger scale dithers. 



#18 NMBob

NMBob

    Surveyor 1

  • *****
  • Posts: 1,608
  • Joined: 21 Nov 2015

Posted 02 June 2017 - 11:47 AM

To complement Jon's observation;

Lot's of 30sec subs from a severe LP zone with no true dark nights anymore,

QHY 163M 280x30 sec L, 90x30secRGB/channel, GSO RC 8, AP 0.67 red, Avalon Lineair

 

NGC7023 by Yves, on Flickr

VERY nice! We should quit goofing around and GO there. :)

 

So, now, what would be the change in, well, everything, if you were to try and make a picture something along these lines with a color camera, like a ASI1600MC-cooled? I just have an ASI174MM and a filter wheel, but I'm just not sure I would have the patience to do LRGB, and I don't expect to be producing anything this nice, either. I'm just trying to gauge how much worse you could do with a color camera. I'm just trying to get an idea of what you lose (other than total control) by not using a mono camera that you can't get back with a color one. You know what I mean? :)

 

I see all of these great pictures meticulously made with mono cameras. I'm not sure I could do that. Hmmm...I want to wow the relatives, and not you guys. Maybe that's the quality I'm shooting for at this point. :)

 

Thanks!



#19 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 02 June 2017 - 12:11 PM

 

 

from nebulosity help file

suvowmner...that is the way I use to do it until I read the thread on this camera regarding short exposure times. I managed to get a few hours of data on M101 the other night at 3 min subs. They came out good, but I do not have enough of them as they are still a bit grainy. I have 69x3mins usable subs. They look good, but I am estimating I need another ~80 subs. The BG median was around 1280. So, I think that thread might produce good results in dark skies with this camera, but I think it is woefully inadequate in light polluted skies.

 

Kaos

 

So with a 16-bit median around 1280, that is a 12-bit median of around 80, minus the 10 ADU bias offset so 70 ADU of true signal. At a Gain setting of 0, that is 4.88e-/ADU, which means your median background signal was 342e-. That is a darn good background signal. You'll handily swamp the read noise with that, regardless of what rule you might want to follow. With the 20x rule you would need a background signal of 70e-, and with the 3x rule you would need a background signal of only 37e-. So, fundamentally, there is no NEED to expose longer...not from an IQ standpoint. The only reason to expose longer is just to stack fewer subs...however, that is not necessarily ideal for this camera. You really want to stack 100+ subs, well dithered subs at that, to get the best results.

 

I recently imaged a group of galaxies, with 30 second subs on my f/4 scope under very bright skies at Gain 76 with an offset of 15. This gain has 2e- read noise and an ~8192e- full well. My backgrounds measured ~2600 16-bit, which is 162 12-bit. Subtracting the 15 ADU offset, I had 148 ADU, which at 2e-/ADU is 296e-. Again, I've more than sufficiently swamped the read noise, regardless of what mathematical rule you use to determine what is enough to swamp. I stacked 413 of these 30 second L subs, dithered aggressively every 4 subs, and ended up with this:

 

 

 

Here is a closeup of NGC5371, to show the noise quality:

 

 

 

I think the noise looks great, which BTW is as it was in the original stack, this image so far has had zero noise reduction of any kind, only stacking and gradient removal. This camera delivers wonderful results when you stack a lot of frames. The effective real-world bit depth here is 16.4 bits. 

 

Now, the stretch is a fairly hefty one. I plate solved and annotated the image, and highlighted all the background fuzzies between 18.5 and 19.5 mag surface brightness:

 

 

 

The faintest object highlighted in this field is 19.49mag. The faintest object found in the entire full field of the image was over 21mag! My skyfog levels measured with an SQM-L were 18.1 and 18.3 on the nights I imaged.

 

Jon...here is a cropped version of the progress on my  M101 project. It is 69 x 3-min subs taken @ F7. I dithered after each frame (probably too much). It is calibrated and there was minimal stretching done on this image. As indicated, the BG on most subs came in at ~1280 with a high of around 1390 as it got lower toward the west.

 

 

 

I think some of the faint arms may fill in a little better with more data (perhaps 3-4 hours). I  do not want to have to stretch it a lot because some of the brighter stars will begin to really blow out which I think would ruin the image.The three small galaxies visible in this image should also be much more visible. I am also not really happy with the noise level apparent in the target with this number of subs. The .jpeg hid some of the grainy stuff visible in the target. In your opinion, how many more subs do you think it would take to eliminate the graininess visible? Also, how much per color channel do you think would be enough. I was thinking minimum of 24 3 min subs per channel.

 

Kaos

Attached Thumbnails

  • M101-L-Cropped.jpg

Edited by Kaos, 02 June 2017 - 12:14 PM.


#20 Kaos

Kaos

    Messenger

  • -----
  • topic starter
  • Posts: 409
  • Joined: 02 Dec 2015

Posted 02 June 2017 - 12:20 PM

Here is a closeup of the center...the grainy effect is much worse in .TIFF and FITS.

 

 

Attached Thumbnails

  • Zoom M101-L Cropped.jpg


#21 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 02 June 2017 - 01:33 PM

I am honestly not sure what you are expecting. Do you have some reference example you are using that is making you feel your image is too grainy? Lacking your reference point, the data you are sharing looks good for 3h27m of integration in a more heavily light polluted zone. 

 

You are imaging at f/7, which is more on the slow side of an f-ratio. Additionally, the camera noise is NOT your problem...the noise from light pollution is. If you are swamping the read noise by nearly 100x, neither read noise, nor quantization error, nor any other source of camera noise, is your problem. Your problem is just plain and simply that you have high LP. That would be a problem with any camera. In a red zone, to get crisp, clean results at f/7, you'll likely need 10 hours or more of integration. Again, that is purely because of the LP...has nothing to do with the camera noise at all, not when you are swamping the camera noise by such a huge amount. 



#22 Nocturnal

Nocturnal

    Surveyor 1

  • *****
  • Posts: 1,850
  • Joined: 14 Sep 2005

Posted 02 June 2017 - 01:36 PM

 

To complement Jon's observation;

Lot's of 30sec subs from a severe LP zone with no true dark nights anymore,

QHY 163M 280x30 sec L, 90x30secRGB/channel, GSO RC 8, AP 0.67 red, Avalon Lineair

 

NGC7023 by Yves, on Flickr

VERY nice! We should quit goofing around and GO there. smile.gif

 

So, now, what would be the change in, well, everything, if you were to try and make a picture something along these lines with a color camera, like a ASI1600MC-cooled? I just have an ASI174MM and a filter wheel, but I'm just not sure I would have the patience to do LRGB, and I don't expect to be producing anything this nice, either. I'm just trying to gauge how much worse you could do with a color camera. I'm just trying to get an idea of what you lose (other than total control) by not using a mono camera that you can't get back with a color one. You know what I mean? smile.gif

 

I see all of these great pictures meticulously made with mono cameras. I'm not sure I could do that. Hmmm...I want to wow the relatives, and not you guys. Maybe that's the quality I'm shooting for at this point. smile.gif

 

Thanks!

 

 

Why would it be any different? The sensor is the same after all. You can't compare exposure times and total exposure lengths between different scopes and different locations anyway so why try to adjust for the bayer matrix vs RGB filters? You capture in raw and look at the ADU values you get. Noise figures are the same for the color and the mono sensor so whichever advise you like regarding gain settings applies to both. Then pick an exposure setting that works for your target/sky/telescope/mount combo and take as many of these as you can. Ultimately what matters most is total exposure time. Dividing say 6 hours in 30s subs or 90s subs will result in subtle differences you will most likely not notice immediately. Because of the low read noise and super fast download times there is no or very little penalty for taking lots of short exposures and those are much much easier to take than longer ones.



#23 Nocturnal

Nocturnal

    Surveyor 1

  • *****
  • Posts: 1,850
  • Joined: 14 Sep 2005

Posted 02 June 2017 - 01:38 PM

Here is a closeup of the center...the grainy effect is much worse in .TIFF and FITS.

 

Image looks good to me. Be sure to flatten even if your optics don't require it as the varying gain response of the pixels may introduce variation that is not noise.



#24 suvowner

suvowner

    Viking 1

  • *****
  • Posts: 662
  • Joined: 22 May 2016

Posted 02 June 2017 - 01:50 PM

 

from nebulosity help file

suvowmner...that is the way I use to do it until I read the thread on this camera regarding short exposure times. I managed to get a few hours of data on M101 the other night at 3 min subs. They came out good, but I do not have enough of them as they are still a bit grainy. I have 69x3mins usable subs. They look good, but I am estimating I need another ~80 subs. The BG median was around 1280. So, I think that thread might produce good results in dark skies with this camera, but I think it is woefully inadequate in light polluted skies.

 

Kaos

 

Sams test images for that blog post were done using a F/4 and a F/2.8 scope, my experience with the asi 1600 color is with hyper star on a C8, which is F2.1 , I am not sure as to which is better in light pollution short exposure with fast 2-4 F ratio, vs longer exposure and slower F ratio, I am sure Jon could give some insight, simply from the perspective of getting multiple images to stack the fast F ratio sure gets you those 100-300 images a lot quicker......

 

Jon correct me here but I think it is fair to say, mono over color cameras have there biggest advantage in light pollution for narrowband imaging, for lrgb with mono and filters the advantage over a color camera in light pollution is much smaller approaching negligible ???

 

at some point the drive to a dark sky site vs image acquisition and processing time in light pollution which is better time spent gets interesting......no way to escape the principals of signal to noise and light pollution is noise...


Edited by suvowner, 02 June 2017 - 01:54 PM.


#25 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 26,034
  • Joined: 10 Jan 2014

Posted 02 June 2017 - 01:56 PM

 

 

from nebulosity help file

suvowmner...that is the way I use to do it until I read the thread on this camera regarding short exposure times. I managed to get a few hours of data on M101 the other night at 3 min subs. They came out good, but I do not have enough of them as they are still a bit grainy. I have 69x3mins usable subs. They look good, but I am estimating I need another ~80 subs. The BG median was around 1280. So, I think that thread might produce good results in dark skies with this camera, but I think it is woefully inadequate in light polluted skies.

 

Kaos

 

Sams test images for that blog post were done using a F/4 and a F/2.8 scope, my experience with the asi 1600 color is with hyper star on a C8, which is F2.1 , I am not sure as to which is better in light pollution short exposure with fast 2-4 F ratio, vs longer exposure and slower F ratio, I am sure Jon could give some insight, simply from the perspective of getting multiple images to stack the fast F ratio sure gets you those 100-300 images a lot quicker......

 

Jon correct me here but I think it is fair to say, mono over color cameras have there biggest advantage in light pollution for narrowband imaging, for lrgb with mono and filters the advantage over a color camera in light pollution is much smaller approaching negligible ???

 

at some point the drive to a dark sky site vs image acquisition and processing time in light pollution which is better time spent gets interesting......

 

Correct. LRGB+Mono in heavier light pollution greatly diminishes the advantages over OSC/DSLR. The main remaining advantage, based on my experience, is that mono+LRGB (or possibly just mono+RGB) will give you better color. The bandpasses of OSC/DSLR color filters in the CFA are usually pretty sloppy. They have narrow peaks which taper off and usually overlap, often quite a lot, which results in the colors becoming rather impure. This leads to higher color noise, so SNR is not as good. However again, in high light pollution, you won't usually notice the the increase in color noise as much as just the loss of color purity and accuracy. 

 

For narrow band, yes, mono has the distinct advantage in a light polluted zone. Additionally, mono has an edge for LRGB at a dark site, and the darker the site, the more the mono+LRGB will gain the advantage. 




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics